Missing institutions in OpenAlex: possible reasons, implications, and solutions

The advent of open science calls for open data platforms with high data quality. As a fully open catalog of the global research system launched in January 2022, OpenAlex features two main advantages of easy data accessibility and broad data coverage, which has been widely used in quantitative science studies. Remarkably, OpenAlex is adopted as an important data source for Leiden university ranking. However, there is a severe data quality problem of missing institutions in journal article metadata in OpenAlex. This study investigates the possible reasons for the problem and its consequences and solutions by defining three types of institutional information—full institutional information (FII), partially missing institutional information (PMII) and completely missing institutional information (CMII). Our results show that the problem of missing institutions occurs in more than 60% of the journal articles in OpenAlex. The problem is particularly widespread in metadata from the early years and in the social sciences and humanities. Using sub-samples of the data, we further explore the possible reasons for the problem, the risk it might represent for distorted results, and possible solutions to the problem of missing institutions. The aim is to raise the importance of data quality improvements in open resources, and thus to support the responsible use of open resources in quantitative science studies and also in broader contexts.